为细化研究公共交通乘客的换乘行为，基于公交IC卡数据，提出了一种区分换乘模式、公共交通站对及线路的乘客换乘时长阈值计算方法。该方法首先设置换乘空间阈值筛选换乘站点对，然后根据每条公交线路的发车时刻计算相应的候车时间阈值，通过构建四阶段模型计算乘客在换乘站点对内的步行时间，进而产生每个换乘站点对之间不同公交线路的换乘时长阈值。基于该方法，采用武汉市公交IC卡数据，分析乘客的换乘时长阈值及换乘行为特征。结果显示，公交乘客的换乘时长阈值和换乘时长存在明显的空间差异性，不同公共交通站点对及线路的换乘时长阈值与其所在空间位置显著相关;乘客的换乘时长服从对数正态分布;常规公交之间的换乘量（约占换乘总量的70.2%）远超常规公交与轨道交通之间的换乘量（约占换乘总量的29.8%）。基于本方法分析乘客的换乘行为，可避免使用传统的固定阈值判别乘客换乘行为造成的误差，提高辨识结果的准确性。

In order to study the transfer behaviors of public transport passengers in detail, based on public transport IC card data, a method was proposed to calculate the transfer time threshold, which could distinguish the transfer mode, station pairs and transit lines. Firstly, transfer space thresholds were identified to select transfer station pairs; secondly, waiting time thresholds of each lines were calculated based on their corresponding departure time; thirdly, passengers′ transfer walking time thresholds were calculated using four-step method; finally, transfer time thresholds of different transit lines between each transfer station pair could be calculated. Based on the proposed calculation method, this paper used IC card data of Wuhan public transport system to analyze passengers′ transfer time thresholds and transfer behaviors. The results showed that the transfer time thresholds and transfer behaviors were significantly spatial-dependent; the transfer time thresholds between different station pairs or in different transit lines were significantly correlated with their spatial locations; the transfer time was log-normal distributed; the amount of transfer trips between bus-to-bus (accounting for about 70.2%) was great larger than the amount of transfer trips between bus-to-metro and metro-to-bus (accounting for about 29.8%). Based on the method of transfer time threshold proposed in this paper, bias caused by the traditional fixed transfer time threshold can be avoided and the accuracy can be improved when identifying passengers′ transfer behaviors.